ACTG A5221 STRIDE: An international randomized

trial of immediate vs early antiretroviral therapy

(ART) in HIV+ patients treated for tuberculosis
D Havlir, P Ive, M Kendall, A Luetkemeyer, S Swindells, J
Kumwenda, J Rooney, S Qasba, E Hogg, J Andersen, I Sanne
ACTG SITE INVESTIGATORS

Mina Hosseinipour
Umesh Lalloo
Valdilea Veloso
Fatuma Some
N. Kumarasamy
Nesri Padayatchi
Breno Santos
Stewart Reid
James Hakim
Lerato Mohapi
Peter Mugyenyi

Constance Benson
Jorge Sanchez
Javier Lama
Jean William Pape
Fred Sattler
Aida Asmelash
Evans Moko
Frederick Sawe
Mauro Schechter
Thira Sirisanthana
Srikanth Tripathy
Judith Aberg

BACKGROUND

HIV-associated TB is a major cause of

morbidity and mortality globally
ART started prior to completion of TB
therapy reduces mortality1
However, the optimal time to start ART
during TB treatment has not been
established
Clinicians often must decide when to start
ART prior to the confirmation of TB
1 Abdool

Karim, NEJM, 2010

HYPOTHESIS

In patients starting treatment for TB, the

immediate initiation of ART (within 2 weeks)
could reduce mortality and morbidity compared
to the early initiation of ART (8-12 weeks)

STUDY DESIGN

Phase IV, randomized, open-label strategy study

HIV+ adults with confirmed or presumed TB

CD4 <250

Two arms: immediate ART (<2 weeks) vs. early

ART (8-12 weeks)

ART regimen: EFV + TDF/FTC

TB treatment regimen: rifampin based, country

approved

STUDY SCHEMA
IMMEDIATE
ART

PRIMARY
ENDPOINT

TB TREATMENT
ART

HIV+
TB
CD4<250
TB TREATMENT

EARLY ART

ART

24
STUDY WEEK

48

STUDY ENDPOINTS

Primary: all-cause mortality and new AIDSdefining illnesses by 48 weeks

Proportions estimated using the Kaplan-Meier method
Stratified analysis by weighting by the inverse of the
Greenwood's variance in each CD4 stratum

Secondary:
Safety
CD4, HIV RNA changes
TB IRIS1
TB outcomes

1 Required

1 major or 2 minor criteria Meintjes, Lancet ID, 2008

RESULTS: Baseline characteristics

Treatment arm
Immediate
N=405

Early
N=401

All
N=806

275

279

554

Asia

29

23

52

N. America

21

18

39

S. America

80

81

161

48%

45%

46%

Median CD4 cells/mm3

(IQR)

70
(34,146)

82
(40,144)

77
(36, 145)

Median log10 HIV RNA

5.39

5.50

5.43

EVF/TDF/FTC

98%

96%

97%

10 days

70 days

n.a.

Study Site
Africa

Confirmed TB

Median time to ART

RESULTS: Proportion with AIDS/Death

P (95% CI for
difference)

Immediate

Early

All Subjects

12.9%

16.1%

0.45 (-1.8, 8.1)

CD4 <50
cells/mm3

15.5%

26.6%

0.02 (1.5, 20.5)

CD4 50
cells/mm3

11.5%

10.3%

0.67 (-6.7, 4.3)

RESULTS: Proportion with AIDS/Death

P (95% CI for
difference)

Immediate

Early

All Subjects

12.9%

16.1%

0.45 (-1.8, 8.1)

CD4 <50
cells/mm3

15.5%

26.6%

0.02 (1.5, 20.5)

CD4 50
cells/mm3

11.5%

10.3%

0.67 (-6.7, 4.3)

Time-to-New AIDS-Defining Illness or Death

by CD4 Stratum

Primary Endpoint: AIDS

AIDS Illness

Immediate ART

Early ART

Total

(N=26)

(N=37)

(N=63)

Cryptococcal Disease

13

Esophogeal Candidiasis

12

Kaposis Sarcoma

11

Pneumocystis pneumonia

Toxoplasmosis

Cytomegalovirus

Non-TB Mycobacteria

Other

Deaths
Cause

Immediate

Early

Total

(N=31)

(N=37)

(N=68)

14

21

Bacterial infection

10

Cryptococcus

CMV

MAC

Lymphoma

Toxoplasmosis

Non AIDS

10

16

Trauma/suicide/ingestion

Unknown

Tuberculosis
AIDS Related

Grade 3 or 4 Clinical Events or

Laboratory Abnormalities
Event

Immediate

Early

Total

Constitutional

8%

8%

8%

Respiratory

4%

4%

4%

Cardiac/Circulatory

3%

2%

2%

Gastrointestinal

4%

5%

5%

Skin

3%

3%

3%

Neurological

5%

7%

6%

ANC < 750/mm3*

9%

17%

13%

Hemoglobin

7%

5%

6%

Platelets <50,000/mm3 *

<1%

3%

2%

Liver transaminase > 5x ULN

6%

10%

8%

ANY

44%

47%

46%

*P<0.05 for ANC and platelets, all other NS

HIV RNA and CD4 Responses

HIV RNA suppression 74% at 48 weeks

No difference between arms

CD4 change 156 cells: entry to week 48

No difference between arms

Frequency and Predictors of

MTB IRIS

Frequency1
# MTB IRIS
Immediate ART

43 (11%)

Early ART

19 (5%)

Predictor

Hazard
(95% CI)1

P
Value

2.5 (1.4, 4.2)

0.001

HIV RNA Higher 1.8 (1.2, 2.7)

0.007

Immediate ART

Confirmed TB2

1P=0.002

1Cox

3.6 (2.0, 6.6) <0.001

analysis, multivariate analysis

2versus probable/not TB

Summary

Overall, immediate ART did not reduce AIDSdefining illnesses and death compared to early
ART
However, for persons with CD4+ counts< 50/mm3,
immediate ART reduced mortality/AIDS
Grade 3 or 4 toxicities did not differ between arms
No differences in HIV RNA suppression rates
(74%) or CD4 rise between arms
TB IRIS was higher in immediate vs early arms

When to Start ART in TB Building

on previous studies
A5221/ STRIDE

CAMELIA1

SAPIT2

806

660

429

Sites

Africa, Asia, S Am, N

Am

Cambodia

S. Africa

Arms

Imm vs 8-12 wk

Imm vs 8 wk

Early vs 24 wk

Endpt

Death/AIDS <50 CD4

Death

Death

CD4 (IQR)

77 (36,145)

25 (11,56)

150 (77, 254)

1 Blanc,

IAC, 2010 2Abdool Karim, NEJM, 2010

Conclusions

Both immediate and early ART strategies are

safe and do not jeopardize CD4 or viral
suppression rates
In patients with CD4 <50, ART should be started
within 2 weeks- delays increase AIDS/death
TB IRIS is more common in those receiving
immediate ART, but does not increase mortality
Implementation of these findings should be a
high priority in HIV and TB programs and will
require coordination with hospital and outpatient
programs

Acknowledgments

Study patients, families and care providers

Other 5221 STRIDE team members: Fran Aweeka, Eva
Purcelle, Ana Martinez, Travis Behm, Patricia Anthony,
Janet Nicotera, Margaret Mensah-King, Stephanie Warner,
Christina Blanchard, Xingye (Shirley) Wu
Carol Suckow and Lynne Jones, Data Managers
HIV and TB care programs
Veronica Miller, Forum for Collaborative Research
NIAID, Richard Hafner
Members of the DSMB
Gilead Sciences and Merck Pharmaceuticals